F. Madeo and M. A. Schnabel (eds.), Across: Architectural Research through to Practice: 48 th International Conference of the Architectural Science Association 2014, pp. 21 29. 2014, The Architectural Science Association & Genova University Press. FAÇADE DESIGN INOVATIONS AND THE FAÇADE DESIGN PROCESS FOR A SUPER HIGH RISE TOWER IN THE MIDDLE EAST Combining the various façade design requirements ROBERT STEPHENS Inhabit Group Introduction Abstract. This paper considers the development of the façade design for a super high rise tower. The relationship and design process between the various design elements is explored with regards to the special requirements for these towers. Keywords. Super high rise, design efficiency, manufacture, panalisation, building movements, structure, thermal, acoustics, weather tightness The terms super high-rise, mega tall, super tall are frequently used to donate towers which are of unordinary high construction. The unusual height brings special challenges which are found only on these structures. Definitions for super tall categories are given as follows: Mega tall above 600 meters CTBUH Super tall above 300 meters CTBUH According to CBTUH/Emporis there are now over 100 towers that are above 300m: Emporis Table 1: Number of super tall buildings CTBUH Over 600m 3 Over 600m 2 Over 500m 6 Over 500m 6 Over 300m 100 Over 500m 84
22 R. STEPHENS The tower that we will be considering is approximately 520 meters tall, so will be at the higher end of the super tall spectrum and will have specific challenges which are not common for more typical towers. The façade works make up a significant proportion of cost on high rise towers. The façade will make up 15-25% of costs for a high rise and 20-25% of costs for a super tall building. (CBTUH) estimates the cost at 14-27% of the overall construction costs. The reason for the significant proportion on cost can be understood by considering the works under the following categories: Scale 1. Scale of works 2. Technical (performance) requirements The sheer surface area of the building envelope results in significant costs. These costs are controlled economically by using process driven design with repetitive elements. Due to the large amount of repetition and area seemingly small changes in design can result in significant costs and for this reason the design must be carefully monitored Technical Requirements The technical (performance) requirements for the façade are diverse in comparisons to other building elements. The façade must provide visual and aesthetic performance, weather proofing, thermal insulation, acoustic insulation and resist structural forces whilst accommodating movements from the building. These increased design diversity constraints can be gauged by comparisons to say the floors and columns which are required to resist predominantly structural forces. The challenge faced by the design team is to balance the diverse façade constraints whilst considering the manufacturing requirements for repetition to achieve an economical yet refined and robust solution. For super tall towers this challenge is amplified by the greater scale and amount of repetition coupled with the more onerous technical requirements such as higher wind loads, larger building movements etc. The façade is also critical in terms of construction program. Interior fit out works are dependent on closure of the building by the façade and the nature of the façade material and manufacturing process has lead times which need to be planned for. In order to get the most balanced and efficient façade design; from the outset it was considered to marry the technical and manufacturing con-
FAÇADE DESIGN PROCESS ON A SUPER HIGH RISE 23 straints with the architectural requirements of the façade. Solutions were aimed for that addressed multiple technical requirements as opposed to finding solutions for each constraint individually. The architectural concepts that related to the facades were as follows: 1. The building is residential so views from inside to out are most important 2. The building should have a refined appearance with an inherent textual quality that sets the tower apart from other towers Starting with the theme of views, the idea was discussed to have a large picture window of maximum transparency. Typically unitised curtain walling found on towers will have mullion spacing s of between 1.2 to 1.8m. To understand if large glass panes and consequently panels could be used, the team had to understand what drivers led to the typical spacing s. The reason for the typical spacing s was glass size availability, installation and storage of panels and integration with the interior fit out and material cost. With this in mind, the team looked at wide panel options of 1.5m, 3m, and 4.5m options with respect to these constraints as tabulated in Table 2below. Table 2: Wide Panel Options With Respect To Constraints: Module Glass size Installation and storage Integration with interior Material cost availability of panels fit out 1.2-1.5 Yes Easy Possible economic 3 Yes More challenging Possible economic 4.5 Yes Very difficult Possible costly
24 R. STEPHENS The review determined that 4.5 meter wide panels did not appear practical or commercially viable but 3m wide panels were viable provided that installation constraints were accounted for. Interestingly it was found that the amount of material used for the 3m panels was similar / comparable to the typical 1.5m spaced panels. This was because that although the 3m spaced mullions were heavier, they are less frequent on the perimeter; so overall the amount of material was similar. Upon 3m panel selection, the study was then further refined to look at the framing and glass size options. Panel Framing and Glass Size Options of 3m wide Panels The ideal was considered as having 3m wide panels. However, the concern was that the glass would have to be heat strengthened for thermal and structural reasons and that consequential roller waves inherent in the heat strengthening process could cause visual impairment. The roller waves run across the glass width dimension and to minimise visual distortion the glass needs to be aligned with the width dimension along the horizontal. It was found that 3m wide panels (allowing the roller waves to align horizontally) were now available, as was lamination of the same which would also likely be required for structural and acoustic performance reasons. Hence the 3m side panels were confirmed as feasible. This illustrates how quickly the glass
FAÇADE DESIGN PROCESS ON A SUPER HIGH RISE 25 manufacturing industry is moving as 2 years ago fabrication of these size panels on a large scale would not have been commercially feasible. The position of the stack joint was then reviewed (the horizontal joint between panels.) a more economical design can be achieved by positioning the stack joint above the floor. This is because the mullions are designed as continuous across floors and the position of the stack joints has a relieving effect with respect to force and deflection. The savings on design are significant and can be up to a 40% saving of materials according to Inhabits calculations. However, having the stack joint above floor must be considered carefully as it can potentially cause visual impairment. For this project, given the scale it was felt that it was imperative to make this saving and adapt the building architecture to this stack joint alignment. The final constraint to be addressed was transport and installation of the panels. The large panels could be transported by road in open top containers, however manipulation in and around the building would be challenging. The challenge was overcome by utilising A key feature of the towers which is the sky gardens. The sky gardens are open breaks in the structure of the tower. It was considered that the sky gardens could be utilised as lay down areas for the panels and thus alleviate the large panel manipulation and installation concerns. The decision was thus taken to move ahead with the 3m wide panel design.
26 R. STEPHENS Further refinements - Panel zones soft and hard zones. The panel set out must be aligned with the column set out geometry. For the project, the column spacing was not regular and thus in order to get maximum efficiency from the design, the panel set out and size was reviewed by considering some panels as having fixed width, whilst some having soft (variable) widths. Large panels were placed at room centres whist narrower panels were placed at wall, and column locations. This set-out was repeated across the façade to form a rhythmic modulation. Option A Option B Fixed width panels shown in blue It was determined that the area of narrow panels was greater than that of the centre panels. For this reason, variable (soft zones) were chosen for the centre panels with fixed width narrower side panels. Transparency, Solar and Visual Requirements. Current legislation in Dubai and responsible sustainable design requires the use of a Shading Coefficient of 0.23. The Shading Coefficient has a consequent impact with regards to the views. The highest transparency possible with a shading coefficient of 0.23 would be approximately 40%. 40% would not create the sought after light airy ambiance internally. New and Current Dubai government legislation allows the arbitrary performance requirements to be satisfied by demonstration of an equivalent system. In discussions, the design team proposed that an equivalent model could be demonstrated to enable clearer glass to be utilised for the large centre pane locations. The approach was to specify higher light transmission glass which had inferior solar performance for the large centre panes. This inferior performance would be compensated by dark / opaque side panel glazing of much higher solar performance. The solution was considered as an ergonomic design as the adjacent panels were often in front of curtains and adjacent to interior walls. The side panels were less critical in terms of their views as opposed to the
FAÇADE DESIGN PROCESS ON A SUPER HIGH RISE 27 centre panes which were considered as extremely critical. A further benefit would be that from the outside, the opaque panels would minimise / alleviate the banding effect seen on many facades between column spandrel areas and clear areas. This was seen by the architect to add the intrinsic visual quality to the façade which was being aimed for. Thermal Mechanics An energy model was set up to compare a standard construction with a S.C. of 0.23 vs what would be required if adjacent opaque panels were used. The results showed that a S.C. of 0.35 (allowing 50-60% light transmission) could be used with opaque panels with a S.C. of approximately 0.15 as illustrated below: Base building standard 40% visual areas with S.C. of 0.23 Improved design with balancing of vision panels with opaque panels Upon confirmation of the visual and thermal requirements, the make-up of the opaque panels was investigated. Various forms were considered. The preferred choice was to insert reflective metal louvres in between the IGU s. There were a number of reasons that they were preferred which are as follows: 1) the external appearance of the glass would be similar to the base glass, 2) the louvres would still allow oblique views through the glass. 3) The louvres could be designed to sparkle to give an inherent special textuality to the façade.
28 R. STEPHENS Large quantities of light being reflected off the aluminum louvre blades behind the glass Areas of light blue and turquoise indicate high reflections whereas areas in green indicate lower reflection White areas indicate sunlight incident on the backpans The close up indicate the variation in luminance across the facade Area of white and green variance indicate the spark of the building Analysis of sparkle effect Structural Loads, Acoustic & Movement Requirements Building movement is derived from the wind tunnel test. The wind tunnel test measurements took time to prepare, and the design of the building was fluid. Thus the above stages of design were carried out with assumed windload values, acoustic performance values and building movement values based on Inhabits experience of similar towers. Once the actual values were available a further refinement of the design was carried out to ensure that the exact values were accounted for. This is standard practice for even standard towers and must be assessed very carefully for super high rise towers to ensure the design development and project budget development proceeds consistently without disproportional and unforeseen costs. A similar approach is required for the preliminary acoustic design. Design Development - Structure and Acoustics Once the acoustic monitoring and wind tunnel testing had been carried out, Two maps of the building facades were made; one for structural requirements and one for acoustic requirements. Required glass thicknesses / make up s were marked dependent on each requirement. The maps were then superimposed and homogenised to have a consistent set of thicknesses required to meet the acoustic and structural requirements. The end result is a highly efficient glazing design which is where much of the façade cost lies. Building Movement Building movement is of particular concern on super high rise buildings. Miscommunication, misunderstanding have in the past on many occasions led to abortive works and claims, (e.g. Burj Khalifa which had approx. 40 floors of back stitched beams added at considerable expense to stiffen the floors and reduce deflections.
FAÇADE DESIGN PROCESS ON A SUPER HIGH RISE 29 The building deflections that affect the façade design are: 1. Lateral movements: wind, seismic 2. Vertical movements: column shortening, differential shortening and slab edge / perimeter deflections. The lateral movements occur as 3 modes: 1. Drift shear racking of the structure, 2. Moment curvature of the vertical elements of the structure 3. Torsion twisting of the structure which is a form of racking. The moment curvature gives differential compression to the columns and alike which needs to be allowed for along with the vertical movements described above The torsion of the structure gives large racking movements to the facade and is often missed in the design. The large movements are because the façade perimeter is furthest away from the centroid of twisting (usually around the centre core) the small angular rotation thus amplifies to a large perimeter deflection. The vertical movements occur in 3 modes: 1. Slab deflection, 2. column shortening, and 3. Differential shortening. The differential column shortening has a component that translates to a lateral drift which needs to be allowed for as well as the other drift movements. The gravity deflections also occur over 3 different time periods: immediate, intermediate (transient) and long term creep. As can be seen the above modes and time frames result in many permutations of deflection requirement. At the minimum there would be 3x3x3=27; that is assuming a single deflection case for each mode which is never the case. The deflection analysis is thus carried out using spreadsheets to ensure all cases are covered. The deflection analysis determines Joint sizes both vertical and horizontal and bracket design and adjustment and is critical for smooth progression of the design and cost plan. References Watts, S.: CTBUH Journal 2010, The Economics of High-rise CBUTH Tallest building lists www.skyscrapercentre.com Emporis The world s tallest buildings www.emporis.com